Akbar, Dimas Renandra (2026) FORMULATION OF BUSINESS INTELLIGENCE BASED ON MACHINE LEARNING AND LEAN CONSTRUCTION FOR PROJECT CONTROL AT PT ROYAL PROJECT MANAGEMENT. Undergraduate thesis, UPN "Veteran" Jawa Timur.
![[img]](https://repository.upnjatim.ac.id/style/images/fileicons/text.png) |
Text (Cover)
22012010142.-cover.pdf Download (2MB) |
|
Text (Bab 1)
22012010142.-bab1.pdf Download (224kB) |
|
Text (Bab 2)
22012010142.-bab2.pdf Restricted to Repository staff only until 3 July 2029. Download (512kB) |
|
Text (Bab 3)
22012010142.-bab3.pdf Restricted to Repository staff only until 3 July 2029. Download (454kB) |
|
Text (Bab 4)
22012010142.-bab4.pdf Restricted to Repository staff only until 3 July 2029. Download (1MB) |
|
Text (Bab 5)
22012010142.-bab5.pdf Download (156kB) |
|
Text (Lampiran)
22012010142.-lampiran.pdf Restricted to Repository staff only Download (702kB) |
|
Text (Daftar Pustaka)
22012010142.-daftarpustaka.pdf Restricted to Repository staff only until 3 July 2029. Download (181kB) |
Abstract
Construction project control is generally still reactive in nature, causing schedule deviations and cost overruns to often be detected only after delays and budget escalations have occurred. This study aims to formulate a business intelligence framework for weekly project control, develop an XGBoost predictive model as an early warning system, and propose corrective action recommendations based on lean construction principles. A quantitative approach was employed in a Type-45 housing project managed by PT Royal Project Management, utilizing weekly data consisting of project schedules, cost budgets (RAB), WBS-based physical progress, and actual expenditures. Data processing was conducted through earned value management, S-curve analysis, time-based splitting, and walk-forward validation. The results indicate that the integration of Planned Value (PV), Earned Value (EV), and Actual Cost (AC) into the BI dashboard produced consistent performance indicators. Furthermore, the XGBoost model demonstrated superior performance compared to the baseline model, achieving MAPE values of 3.81% for EV prediction and 1.16% for AC prediction. The predictive outputs were subsequently converted into risk status categories, normal, alert, and critical, and linked to lean-based SOPs to accelerate managerial response.
| Item Type: | Thesis (Undergraduate) | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Contributors: |
|
||||||||
| Subjects: | T Technology > T Technology (General) T Technology > T Technology (General) > T58.6-58.62 Management Information Systems |
||||||||
| Divisions: | Faculty of Economic and Business > Departement of Management | ||||||||
| Depositing User: | Dimas Renandra | ||||||||
| Date Deposited: | 03 Jul 2026 03:05 | ||||||||
| Last Modified: | 03 Jul 2026 03:10 | ||||||||
| URI: | https://repository.upnjatim.ac.id/id/eprint/54401 |
Actions (login required)
 |
View Item |